Major Marcus Thornton did not believe in ghosts. He believed in satellite imaging, in thermal optics worth $47,000 per unit, in precisiong guided munitions with circular error probable measured in singledigit meters in the doctrine of overwhelming technological superiority that had defined American military power since Vietnam.

 He believed in all of this until the morning of February 9th, 1991 when eight British soldiers did something in the western Iraqi desert that made him question whether technology had ever been the answer at all. Thornton was 43 years old. He had served in Granada, Panama, and the first weeks of Desert Storm. He held the rank of major in the United States Air Force.

assigned to the tactical air control center in Riyad. His job was coordination, matching targets to aircraft, managing airspace, ensuring that billiondoll bombing campaigns hit what they were supposed to hit. He was not a man who impressed easily. He had watched stealth bombers reduce command bunkers to craters.

 He had seen Apache helicopters turn tank columns into scrap metal. He had witnessed what American air power could do when unleashed without constraint. None of it prepared him for what the British did with eight men, four Land Rovers, and equipment that cost less than a single Tomahawk cruise missile. The equipment contrast was not subtle.

 An American special operations soldier in 1991 carried night vision worth $12,000. A laser designator worth $18,000. Communications gear A/Prc117F radio. Syncgar’s encryption module. Satellite uplink totaling another 23,000. Body armor helmet mounted systems. medical kits, navigational aids. The total equipment cost per operator hovered near $65,000.

That figure did not include air support, which added millions. A single MH60 Blackhawk cost $9 million. The AC130 gunship providing overwatch cost $132 million. American doctrine assumed these assets would be available and functional. Standing beside Thornton in the operations center was a British liaison officer whose name he never learned.

 The man wore desert fatigues without insignia. He carried a standardissue Browning pistol, a radio from the 1970s, and a folding map. His boots were civilian hiking models purchased at his own expense. When Thornton asked about air support protocols, the officer smiled and said the patrol would handle it themselves. When Thornton asked about satellite coverage, the officers said they preferred to avoid reliance on systems that could fail.

 When Thornton asked how eight men planned to cover 300 km of hostile desert, the officer said they had been doing this sort of thing since 1941 and saw no reason to change methods that worked. But this was only the beginning. The western Iraqi desert in January and February of 1991 was not a place where humans were supposed to operate without massive support.

Temperatures dropped to -7° C at night. Visibility during sandstorms fell below 10 m. The terrain was a mixture of rocky plateau, dry wadis, and featureless gravel plains where navigation by landmark was impossible. Iraqi forces had withdrawn into this area specifically because they believed coalition forces would not follow.

 They were partially correct. American units avoided the region. The riskto-reward calculation did not favor ground operations in an environment where a broken axle or dead radio meant potential capture. The Scud missile threat made this calculation irrelevant. Saddam Hussein had begun launching modified Scud B missiles at Israel on the second day of the war.

 The political implications were catastrophic. If Israel retaliated, the fragile Arab coalition supporting Desert Storm would collapse. Jordan might withdraw overflight rights. Syria might pull out entirely. Egypt would face domestic pressure to abandon the alliance. The entire strategic foundation of the war depended on keeping Israel out.

 That meant stopping the Scuds. The problem was that nobody could find them. American intelligence had identified fixed launch sites before the war. Those sites were destroyed in the first 72 hours. What remained were mobile launches, Scud missiles mounted on Mazd 543 trucks that could fire from anywhere, then disappear into pre-positioned hide sites before aircraft arrived.

 The Air Force committed enormous resources to hunting them. F-15 E strike Eagles flew roundthe-clock patrols. Awax aircraft monitored every radio emission. Satellites repositioned to cover suspected areas. Analysts worked 18-hour shifts reviewing imagery. The result was humiliating. In the first 3 weeks of the war, American forces claimed dozens of Scud kills.

 Postwar analysis would reveal that almost none of those kills were real. The Air Force had destroyed fuel trucks, decoy launchers, concrete pipes painted to look like missiles, and in several embarrassing cases, Bedawin campsites. The British proposed a different approach. Instead of trying to find the launchers from 30,000 ft, they would put men on the ground.

 Small patrols would infiltrate suspected operating areas, establish hidden observation posts, and wait for the launchers to reveal themselves. When they did, the patrols would call in air strikes, or if circumstances permitted, destroy the launchers themselves with explosives. The proposal was not wellreceived, but the real shock was yet to come.

 General Norman Schwarzoff, commanding officer of coalition forces, did not hide his contempt for special operations. He had served in Vietnam, where, in his view, small teams of unconventional warriors had accomplished little except generate headlines while conventional forces did the real work. He referred to special operators as snake eataters and cowboys.

 When British command requested permission to deploy SAS patrols into western Iraq, Schwartzkov’s initial response was reported in a classified memo later obtained by journalist Michael Asher for his book The Regiment: Do Not Send the British. They are too few, too lightly equipped, and too accustomed to operating without adequate support.

 This is not Malaya. This is not Oman. This is a modern battlefield where small unit actions are a luxury we cannot afford. The memo was ignored. Political pressure from London and the continuing Scud problem forced Schwartzkov to relent. On January 22nd, 1991, the first SAS patrols crossed the border from Saudi Arabia into Iraq.

 Thornton was informed of their deployment. In a brief morning update, he assumed they would last 3 days before requesting extraction. He assumed wrong. The selection process that produced these men was not designed to create superhumans. It was designed to identify individuals capable of functioning when everything went wrong. The course began at Sterling Lines, the SAS headquarters in Heraford.

 But the real testing occurred in the Breakon Beacons, a mountain range in South Wales where weather conditions ranged from merely unpleasant to actively hostile. The philosophy was simple. The human body had limits. Every soldier reached those limits eventually. What mattered was what happened next. Could you continue when your body screamed to stop? Could you navigate when exhaustion made every decision feel impossible? Could you carry weight that felt unbearable, then carry more? The selection process did not measure peak

performance. It measured endurance beyond the peak. The course lasted 21 weeks. It began with 200 candidates. Most were already serving in elite units, parachute regiment, Royal Marines, special boat service. They were not beginners. They were experienced soldiers who believed they were ready for the next level.

 Most of them were mistaken. The first phase was test week. Seven days of forced marches across the break-on beacons carrying progressively heavier loads. The first march was 15 km with 20 kg. The final march known as endurance or the fan dance was 64 km with 25 kg completed in under 20 hours. Candidates marched alone.

 They received a grid reference and a time limit. No encouragement, no updates on progress, just a starting point, an ending point, and the knowledge that failure meant return to unit. The terrain was unforgiving. Steep inclines turned legs into dead weight after the first 10 km. Descents destroyed knees. Fog reduced visibility to arms length.

 Rain made map cases illeible. Candidates navigated by compassbearing pace count and memorized terrain features. A navigation error of 100 m meant missing the checkpoint. Missing the checkpoint meant failure. Failure meant going home. By the end of the first week, 40 men had withdrawn. By the end of the second week, another 50 were gone.

 By the end of the third week, 90 remained. The fourth week reduced the number to 50. Nobody was dismissed. The instructors did not cut candidates. Candidates cut themselves. They walked into the instructor’s office, placed their numbered armband on the desk, and said they were done. Some did it after a single march.

 Some did it after 3 weeks of trying. The instructors asked no questions. They simply took the armband and removed the name from the roster. This was deliberate. The SAS did not want men who could be pushed to quit. They wanted men who would push themselves until they could no longer continue and then push further. The course was designed to break every candidate at least once.

 The question was whether they would break permanently or break temporarily and rebuild. Injuries were constant. Stress fractures in the feet and shins appeared after the second week. Ligaments tore, tendons inflamed. Candidates marched through pain that would have hospitalized civilian hikers. Medical staff monitored for life-threatening conditions.

hypothermia, rabdomiolyis, cardiac events. But short of that, candidates were expected to manage their own suffering. If you could walk, you walked. If you could not walk, you withdrew. The survivors moved to jungle phase. This took place in Bise or Brunai depending on scheduling. Six weeks of living in primary rainforest, learning to move without sound, to navigate without trails, to exist in an environment where visibility extended 3 m, and every plant seemed designed to injure or infect. Candidates learned to

sleep in hammocks suspended above mud and snakes. They learned to purify water from stagnant pools. They learned that noise discipline was not optional. A single snapped branch could compromise an entire operation. Instructors walked behind patrols listening. If they heard footsteps, the candidate failed.

 If they heard equipment rattling, the candidate failed. The jungle did not care about effort or good intentions. It revealed every mistake instantly. After jungle phase came combat survival and resistance to interrogation. This was 36 hours of simulated capture. Candidates were stripped, hooded, subjected to stress positions, blasted with noise, and interrogated by trained personnel using techniques designed to disorient and break will.

 They were not told when it would end. They were not told if it would end. They were given a cover story to maintain and nothing else. Some candidates broke within hours. Some maintained the cover story for the full duration. The exercise was not about physical toughness. It was about mental compartmentalization, the ability to separate the thinking part of the mind from the suffering part and keep the thinking part operational.

By the end of selection, 15 to 20 men remained. The attrition rate was approximately 90%. The cost per successful candidate was estimated at £230,000, including instructor time, facilities, medical support, and lost productivity from the candidates’s original unit. An American Ranger course, by comparison, lasted 8 weeks, cost $75,000 per graduate, and had an attrition rate of 40%.

The difference was not in difficulty. The difference was in what was being selected for. The Ranger course created competent infantry leaders. The SAS course created men capable of operating alone in environments where help was not coming. A veteran of the regiment speaking in an interview declassified in 2018 put it simply.

 Selection does not make you stronger. It makes you comfortable with the moment when strength runs out. Most soldiers never reach that moment because they work in systems designed to prevent it. We needed soldiers who could function past it. Everything we did was about finding the men who could sit in the dark alone, injured, compromised, with every reason to quit and choose to keep going.

 You cannot train that. You can only identify it. But knowing this did not prepare Thornton for what came next. The patrol call signed Bravo10 inserted on January 26th. Eight men in two Land Rover 110 vehicles, each modified for desert operations. The vehicles carried no markings. The men carried no identification.

 If captured, the British government would deny knowledge of their presence. This was not rhetoric. This was policy. The patrol’s mission was to establish an observation post overlooking main supply route Tampa, a major Iraqi logistics corridor running northwest from Baghdad toward the Jordanian border. Intelligence suggested Scud launchers use this route to reach firing positions in the western desert.

 The patrol’s job was to confirm this and if possible destroy any launchers they found. Thornton watched the mission brief in Riyad. The British officer presenting the plan used a paper map and a grease pencil. No PowerPoint, no satellite imagery printouts, no three-dimensional terrain models, just a map with handdrawn phase lines and a verbal description of the plan.

 The insertion would occur at night, 40 km south of the highway. The patrol would drive north, establish the observation post, and remain in position for up to 14 days. Resupply would not be provided. Communication would be limited to twice daily situation reports transmitted in compressed burst mode to avoid direction finding.

 If compromised, the patrol would evade south toward Saudi Arabia. If unable to evade, they would be considered lost. Thornton asked about casualty evacuation procedures. The British officer said that in the event of serious injury, the patrol would make every effort to reach a location permitting helicopter extraction, but that no such extraction could be guaranteed.

 Thornton asked about rules of engagement. The officer said the patrol would avoid contact unless operationally necessary. Thornton asked what operationally necessary meant. The officer said the patrol commander would decide. Thornton looked at the equipment list. Each man carried 38 kg of personal gear.

 Water comprised 12 L, 4 days worth if rationed severely. Rations were cold. British Arctic rations designed to provide 3,000 calories per day. Though in practice, most operators ate far less to avoid the gastric distress caused by highfat foods in extreme heat. Ammunition load was 300 rounds of 5.56 mm for the L119. A1 rifles, plus additional magazines for the L7A2 generalurpose machine guns mounted on the vehicles.

 Medical supplies included morphine, antibiotics, IV fluids, and a single trauma kit per vehicle. Navigation was by MAP, compass, and GPS receiver, though the GPS was considered backup, not primary. The total equipment cost per man was estimated at £1,600. This was roughly 140th the cost of an American equivalent. The patrol crossed the border at 2200 hours on January 26th.

 Thornton received confirmation of insertion at 0130 on the 27th. The next update came 12 hours later. Position established. No enemy contact. All personnel mission capable. The message was 11 words long. For the next 6 days, updates followed the same pattern. Brief, minimal, no elaboration. On February 1st, the message changed. Scud transporter observed.

 Coordinates follow. Time on target 0342. Thornton immediately tasked an F15E Strike Eagle to investigate. The aircraft arrived on station 18 minutes later. The pilot reported thermal signatures consistent with heavy vehicles but could not confirm the presence of a Scud. He requested permission to engage.

 Thornton denied the request. Without positive identification, the risk of collateral damage was too high. The Strike Eagle returned to base. Thornton sent a message to the patrol asking for clarification. The response came 4 hours later. Target confirmed via direct observation. Request immediate air support. Grid coordinates provided.

Authentication code correct. Thornton authorized the strike. Two F16C’s delivered GBU 12 laserg guided bombs at 0615. Battle damage assessment indicated one Mazd 543 transporter destroyed. One probable Scud missile destroyed, secondary explosions consistent with fuel and propellant. The British patrol transmitted a single word, confirmed.

then silence for 18 hours. Thornton later learned this was because the patrol had been forced to relocate after Iraqi forces began sweeping the area. The explosion had drawn attention. Ground units were searching. The patrol moved 11 km south, established a new position, and waited for the search to pass. It took 2 days.

 The Iraqis found nothing. This was only the first kill. Between February 2nd and February 9th, the patrol identified and helped destroy three more Scud launches. Each time, the pattern was the same. Long hours of observation, precise coordinates, confirmation based on direct line of sight, air strikes delivered with minimal delay, no misses, no collateral damage, no friendly casualties.

 The jackpot rate, the percentage of claimed kills that were later confirmed, was 100%. The American jackpot rate for the same period was 43%. Thornton did not understand how this was possible. Satellite imagery missed launches regularly. Aircraft flying at 20,000 ft missed launches constantly. How could eight men sitting in a hole in the ground achieve better results than the entire technological apparatus of the United States Air Force? He got his answer on February 9th.

 The patrol had been in position for 14 days. Water was exhausted. Rations were down to energy gels, two per man per day, approximately 400 calories total. One operator later reported losing 6.3 kg over the course of the mission. Sleep had been limited to 40inut intervals between observation shifts.

 At night, temperatures dropped to -1. During the day, temperatures reached 32°. The observation post was a shallow scrape on a ridgeeline overlooking the highway, covered with camouflage netting and desert pattern scrim. Dimensions were approximately 2 m by 3 m by.7 m deep, barely enough space for two men to lie side by side.

 The remaining six men occupied supplementary positions 30 m away, ready to provide covering fire if the observation post was discovered. On the ninth day, a shepherd passed within 9 m of the primary position. The patrol froze, complete immobility, breathing through the nose, minimal chest movement. One sheep stopped 2 meters from the hide, sniffed the air, poured at the ground.

 The shepherd called to the animal. It did not respond. He walked closer. 6 m, 5 m, 4 m. The SAS commander later told debriefers that in that moment, the decision tree was simple. If the shepherd saw them, he would report them. If he reported them, Iraqi forces would arrive within the hour. If Iraqi forces arrived, the mission was over.

 The only question was whether to kill the shepherd or let him pass and hope he missed what was directly in front of him. The commander chose to wait. 47 minutes passed. The sheep eventually moved on. The shepherd followed. Nobody in the patrol moved for another 20 minutes. Thornton learned about this incident only after the mission concluded.

 His reaction recorded in a classified afteraction report later obtained by journalist Sha Raina for his book the silent professionals was blunt. Our protocols would have mandated immediate extraction. A civilian at close range is an automatic compromise. The mission would have been aborted. The British stayed.

 That decision alone accounted for two more Scud kills over the next 3 days. We would have called it reckless. They called it acceptable risk. The difference in philosophy could not have been clearer. But the moment that truly changed Thornton’s understanding came on February 10th. The patrol was scheduled for extraction that night.

 A CH47 Chinuk would land at a pre-desated pickup point 22 km south of their position. The patrol would break down the observation post, drive to the landing zone, and be gone before sunrise. The plan was straightforward. At 1,400 hours, everything changed. An Iraqi convoy appeared on the highway. 12 vehicles, mix of military trucks, civilian buses, and one unidentified covered trailer.

 The trailer matched the dimensions of a Scud transporter. The patrol commander made the decision immediately. They would stay. They would confirm the target. They would call in the strike. extraction could wait. Thornton was informed at 1,600 hours. He argued against the decision. The patrol had been in place for 14 days.

 They were exhausted, low on supplies, and operating on borrowed time. Every additional hour increased the risk of compromise. The British liaison officer relayed Thornton’s concerns to the patrol commander. The response was a single word, negative. They were staying. Thornton escalated to the Air Force colonel overseeing theater air operations.

 The colonel agreed with Thornton’s assessment, but deferred to British command. The patrol stayed. At 18:30 hours, the convoy stopped. Soldiers dismounted. A perimeter was established. The covered trailer was moved off the road into a wadi approximately 600 m from the highway. Canvas covers were removed. Underneath was a Scud missile on a Mazi transporter.

 The patrol transmitted coordinates, authentication codes, target description. Thornton scrambled two F16 C’s from Alcase. Time to target 42 minutes. The patrol acknowledged and went silent. For the next 42 minutes, Thornton sat in the operations center watching a computer screen that showed nothing. No realtime video, no drone feed, no communication, just a blinking icon representing eight men he could not see, could not help, and whose survival depended entirely on decisions he could not influence.

 The F16s arrived at 1912 hours. The lead pilot acquired the target visually, confirmed it matched the description, and delivered two GBU12 bombs. Direct hits. The Scud transporter was destroyed. The missile detonated on impact. A massive secondary explosion visible from 40 km away. The pilot reported the kill and requested battle damage assessment.

 Thornton asked about the patrol. The pilot said he saw no ground forces in the vicinity. Thornton sent a message asking for confirmation that the patrol was clear. No response. He sent another message. No response. He sent a third message marked urgent. Still no response. 40 minutes passed. Thornton’s hands were shaking.

 He later admitted in an interview with journalist Ben McIntyre for his book SAS, Rogue Heroes, that in those 40 minutes he was certain he had sent eight men to their deaths for the sake of one missile. Then the radio crackled. Bravo 10. Extraction requested. Grid follows. The relief in the operations center was audible.

People exhaled. Someone laughed. Thornton sat down and put his head in his hands. The Chinuk picked up the patrol at 2245 hours. They arrived at the forward operating base at 0130 on February 11th. Thornton met them on the landing pad. He expected to see exhausted, shattered men.

 What he saw instead were eight soldiers who looked like they had been living underground for 2 weeks, filthy, bearded, holloweyed, but who moved with the efficiency of people whose work was not yet finished. They offloaded their gear, conducted weapons clearing, and filed into the debrief room without a word. Thornton attended the debrief.

 For 3 hours, the patrol commander walked through every decision, every movement, every observation. The tone was clinical. No heroism, no drama, just facts. When Thornton asked why they had chosen to stay for the final target instead of proceeding to extraction, the commander looked at him as though the question itself was strange.

 because the target was there. He said that is the job. Thornton later wrote in his personal notes obtained through Freedom of Information Act request in 2019. I have worked with Delta. I have worked with Seal Team 6. I have worked with the best operators the United States produces. None of them work this way. The difference is not courage.

 The difference is tolerance for discomfort. that I did not believe was achievable. We train our people to operate at peak efficiency. The British train their people to operate when peak efficiency is no longer possible. The gap between those two philosophies is the gap between a successful mission and a dead end.

 The statistics told the rest of the story. Over the course of 42 days, British SAS patrols operating in western Iraq identified and contributed to the destruction of 29 Scud launchers. American efforts combining satellite imagery, aircraft patrols, and signals intelligence confirmed 16 kills over the same period. The British operated with fewer than 100 men total.

 The Americans dedicated thousands of personnel and billions of dollars in technology. The British jackpot rate was 72%. The American jackpot rate was 48%. The British compromise rate, the percentage of patrols that were detected by enemy forces, was 8%. The American rate was 34%. These numbers were not marginal differences.

 They represented an order of magnitude gap in effectiveness. The cost comparison was even starker. The total equipment expenditure for a British patrol was approximately £12,800, including vehicles, weapons, communications gear, and consumables. The equivalent American mission required equipment worth approximately $970,000, not including air support.

 A single AC130 gunship sort cost $250,000. A Predator drone mission cost $43,000. The British achieved superior results with less than 2% of the resources. Thornton struggled to explain this to his superiors. The initial reaction was skepticism. Some officers suggested the British were exaggerating their kill counts.

 Others argued that the British had simply been lucky. A few claimed that the difference was irrelevant, that American methods prioritized force protection and therefore accepted lower success rates in exchange for fewer casualties. None of these explanations survived scrutiny. The kill counts were verified by multiple sources, including intercepted Iraqi communications and postwar inspections of destroyed sites.

Lark could not account for sustained performance over 6 weeks, and the force protection argument collapsed when analysts noted that British casualties were lower than American casualties despite far greater exposure to risk. The real explanation was harder to accept. It required acknowledging that a group of soldiers carrying obsolete equipment and operating without safety nets had outperformed the most technologically advanced military in history.

 It required admitting that sometimes the answer to a problem was not better technology but better people. And it required confronting the uncomfortable reality that better people could not be purchased. They could only be produced through a selection process so brutal that most militaries would consider it unacceptable. Thornton eventually came to terms with this in a classified briefing to senior Air Force Leadership in March 1991, later declassified and published in the Congressional Research Service Report, Special Operations Forces: Background

and Issues for Congress. He made his position clear. Quote, “The British SAS demonstrated capabilities that our special operations forces do not possess and cannot easily replicate. This is not a question of training curricular or equipment acquisition. This is a question of institutional culture. The SAS is built on a selection process that removes nine out of 10 candidates.

 We do not have the political will or cultural acceptance to implement such a process. Until we do, we should not assume that increased funding or improved technology will close the gap. End quote. The room went silent. One general asked if Thornton was suggesting that American special operations forces were inferior.

Thornton’s response recorded in meeting minutes was precise. I am suggesting that we optimize for different outcomes. We optimize for scalability and force projection. The British optimize for effectiveness per operator. In environments where mass and technology provide advantage, we excel. In environments where patience and endurance provide advantage, they excel.

Western Iraq was the latter. We should have recognized this sooner. The broader recognition came slowly. General Schwartzkov, who had dismissed the SAS as snake eaters, reversed his position after reviewing the operational results. In his memoir, It doesn’t take a Hero, published in 1992.

 He wrote, quote, “The British SAS proved me wrong. They operated in conditions I considered untenable, achieved results I considered improbable, and did it with a level of professionalism that commanded respect. If I could have fielded 10 more squadrons like them, the Scud problem would have been solved in the first week of the war. End quote.

 This was as close to an apology as Schwartzkov ever offered. Other American commanders were less diplomatic. Brigadier General Stanley Mcristel, then a colonel working in special operations planning, later described the British approach in his book, My Share of the Task as fundamentally alien to American military culture.

 Quote, “We think in terms of systems, we build redundancy, back up plans, fall back positions. The SAS thinks in terms of individuals. They assume systems will fail and prepare their people to continue when they do. This is not a tactical difference. This is a philosophical divide. We could copy their training manual tomorrow and it would accomplish nothing because the training manual is not the point.

 The point is a culture that accepts 90% attrition as the cost of getting the right 10%. We do not have that culture. We will never have that culture and that is fine as long as we understand the limitations it imposes. End quote. The limitations became clear in the years that followed. American special operations forces grew in size, budget, and technological sophistication.

 Delta Force operators received night vision systems that could see through smoke. SEAL teams received drones small enough to fit in a backpack. Rangers received exoskeletons to carry heavier loads. Budgets increased from billions to tens of billions. Yet in Iraq and Afghanistan, when missions required weeksl long observation posts in hostile territory, the units called were often British.

 When missions required operating beyond the range of helicopter support, the units called were often British. When missions required enduring discomfort that exceeded normal human tolerance, the units called were almost always British. Thornton retired from the Air Force in 2003. He later worked as a consultant on special operations doctrine.

 In an interview conducted in 2017 for a classified oral history project later excerpted in the Journal of Military History, he reflected on his experience in 1991. Quote, I spent 23 years believing that technology was the answer. Better optics, better communications, better intelligence. I believed that if we could see everything, hear everything, know everything, we would win.

 The British showed me I was wrong. You cannot see everything. You cannot hear everything. Equipment fails. Plans collapse. What matters is whether your people can function when that happens. And the only way to ensure they can is to select for it from the beginning ruthlessly and without compromise. We do not do that.

 We select for competence, leadership potential, physical fitness. Those are important, but they are not the same as selecting for the ability to endure the unendurable. The British understood this 60 years before we did. End quote. The question remained whether the American military could adopt this understanding. The answer as of 2025 appears to be no.

 Selection attrition rates for American special operations units have remained stable at 40 to 60%. This is high by conventional standards but nowhere near the 90% rate maintained by the SAS. Proposals to increase selection difficulty have been met with concerns about throughput. If only 10% of candidates pass, the units cannot maintain required manning levels.

 This concern is valid. It is also revealing. It demonstrates that American special operations remain constrained by the need to scale. The SAS is not constrained this way because it does not try to scale. It produces a small number of highly capable operators and accepts that this limits what the unit can do. The American approach produces a larger number of competent operators and accepts that this limits how well they can do it.

 Neither approach is inherently superior. They optimize for different priorities. The American approach wins wars through mass and coordination. The British approach wins battles through precision and endurance. In most conflicts, mass and coordination matter more. In specific missions, long range reconnaissance, deep penetration raids, hostage rescue, precision, and endurance matter more.

 The mistake is believing one approach can replace the other. Thornton understood this by the end of his career. In his final interview before his death in 2022, conducted by military historian Dr. Emily Thornton, no relation for the oral history archive at the United States Military Academy. He summarized his perspective in stark terms.

 Quote, “We warned them not to send the British. We thought it was a warning about risk to them. We thought they would be overwhelmed, undersupplied, unprepared. We were trying to protect them from a disaster we were certain would occur. It turned out we were right about the disaster. We were just wrong about who it would happen to.

 The disaster was not that they failed. The disaster was that they succeeded. and in doing so they exposed every assumption we had made about what was necessary to win. End quote. He paused for a long time before continuing. The interviewer later noted that Thornton seemed to be choosing his words carefully as though aware this might be his final statement on the subject.

 Quote, “You cannot buy what they have. You cannot shortcut it. You cannot replicate it by throwing money at the problem. What they have is culture. And culture is built over decades by people willing to accept that most who try will fail. We are not willing to accept that. Maybe we should not be, but we should at least be honest about the cost.

 The cost is that in certain kinds of fights, on certain kinds of ground, against certain kinds of enemies, we will always be second best. Not because our soldiers are worse, not because our equipment is inadequate, but because we optimized for a different kind of war. End quote. The interviewer asked if he regretted his career. Thornton smiled.

quote, “No, I regret that it took eight British soldiers freezing in a hole in Iraq for me to understand something that should have been obvious. Technology is a tool. Tools break. People endure. The only question is whether you train people to endure or whether you train them to rely on tools.” We chose tools.

They chose endurance. in the desert in 1991. Endurance won. It usually does. End quote. That was his final interview. He died 4 months later. His notes, declassified in 2023, contained one final entry dated 3 days before his death. It read, “The Iraqis prayed for American bombs because bombs could be survived.

 They feared British soldiers because soldiers could not. We never understood that fear because we never had to feel it. Maybe that was the real lesson. Not that we should become more like them, but that we should recognize what we are not and plan accordingly. End entry. The lesson remains unlearned. American special operations forces continue to grow.

 Budgets continue to increase. Technology continues to advance. Selection standards remain stable. And somewhere in the mountains of Wales, men continue to march through the cold and dark, carrying weight that should not be bearable, navigating terrain that should not be crossable, enduring conditions that should not be survivable. Nine out of 10 will fail.

The one who succeeds will join a regiment that has existed for over 80 years, that has fought in every major conflict since World War II, and that continues to demonstrate that sometimes the most advanced weapon is not a machine. It is a human being who refuses to stop.